相关参考文献
注意:仅列出部分参考文献,下载原文获取全部文献信息。
Article
Engineering, Industrial
Taotao Zhou et al.
Summary: This paper proposes a reliable fault diagnosis method based on deep learning, which improves the reliability and accuracy of deep learning-based fault diagnosis in safety-critical applications by quantifying predictive uncertainty and introducing a predictive risk-aware strategy.
RELIABILITY ENGINEERING & SYSTEM SAFETY
(2023)
Article
Computer Science, Artificial Intelligence
Juan E. Arco et al.
Summary: The use of automatic systems based on artificial intelligence for medical image classification has greatly improved the diagnosis of various diseases. This study proposes a multi-level ensemble classification system based on Bayesian Deep Learning to maximize performance while also providing uncertainty of each classification decision. The system demonstrates high accuracy in different real scenarios, making it an applicable tool for clinicians.
INFORMATION FUSION
(2023)
Article
Computer Science, Artificial Intelligence
Jian Lin et al.
Summary: This paper proposes a generalized model-agnostic meta-learning method for few-shot fault diagnosis of bearings driven by heterogeneous signals under various operating conditions. The proposed method utilizes a channel interaction feature encoder with multi-kernel efficient channel attention to focus on mutual fault information and effectively extract general diagnostic knowledge for multiple diagnostic meta-tasks. Additionally, a flexible weight guidance factor is designed to adjust the training strategy and optimize the inner loop weights for different diagnostic meta-tasks, improving the overall generalization performance. The method is validated through the analysis of acceleration and acoustic signals of bearings, demonstrating its extensiveness and effectiveness in various few-shot cross-domain scenarios.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Engineering, Mechanical
Xingkai Chen et al.
Summary: Most of the existing research on unsupervised cross-domain intelligent fault diagnosis focuses on single-source domain adaptation, lacking the ability to utilize multiple source domains with diverse diagnostic information. This paper proposes a dual adversarial guided unsupervised multi-domain adaptation network (DAG-MDAN) to better extract common features and integrate multi-source domain knowledge. The DAG-MDAN includes an edge adversarial module (EA-Module) and an inner adversarial module (IA-Module) to enhance domain confusion and a multi-subnet collaborative decision module (MCD-Module) for better fusion decisions.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2023)
Article
Engineering, Industrial
Shen Yan et al.
Summary: This paper proposes a new machinery anomaly detection method called full graph dynamic autoencoder (FGDAE) for complex operating conditions. It develops a full connected graph (FCG) to obtain global structure information and constructs a graph adaptive autoencoder (GAAE) model to aggregate multi-perspective feature information between channels. The method achieves better performance compared to other popular anomaly detection methods on machinery datasets.
RELIABILITY ENGINEERING & SYSTEM SAFETY
(2023)
Article
Engineering, Mechanical
Zuogang Shang et al.
Summary: Deep learning is becoming popular for fault diagnosis, but the lack of explainability and the extraction of weak fault features from noisy signals are the major challenges. To address these limitations, we propose the SPINN framework that embeds signal processing knowledge into deep learning models. We also develop a denoising fault-aware wavelet network (DFAWNet) as a practical implementation of SPINN, which significantly improves diagnostic performance compared to other explainable or denoising deep learning networks.
CHINESE JOURNAL OF MECHANICAL ENGINEERING
(2023)
Article
Automation & Control Systems
Mingxuan Liang et al.
Summary: A probabilistic fault diagnosis framework using a Gaussian process classifier was developed to account for uncertainties in prediction, showcasing unique capability and outperforming other machine learning models in accuracy and robustness in systematic case studies. Sensor fusion with spatial vibration measurements further enhances fault diagnosis performance. Leveraging the probabilistic feature facilitates future research on extended fault diagnosis using limited fault labels.
INTERNATIONAL JOURNAL OF ADVANCED MANUFACTURING TECHNOLOGY
(2022)
Article
Engineering, Mechanical
Yifei Ding et al.
Summary: This paper introduces a novel time-frequency Transformer model to address the shortcomings of traditional models, demonstrating superior fault diagnosis performance in comparison with benchmark models and other state-of-the-art methods.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2022)
Article
Engineering, Mechanical
Dong Wang et al.
Summary: This paper presents a fully interpretable neural network for machine condition monitoring, which combines signal processing and physical feature extraction to automatically localize informative frequency bands.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2022)
Article
Automation & Control Systems
Tianfu Li et al.
Summary: The article introduces a novel wavelet-driven deep neural network, WaveletKernelNet (WKN), for mechanical fault diagnosis, which is more effective than traditional CNN in terms of accuracy and convergence speed. The network is designed with a continuous wavelet convolutional layer to discover more meaningful kernels and provides a customized kernel bank for extracting defect-related impact components.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2022)
Article
Automation & Control Systems
Yiming Xiao et al.
Summary: This article proposes a novel joint transfer network for unsupervised bearing fault diagnosis from the simulation domain to the experimental domain. It uses bearing simulation data to construct the source domain, an improved loss function to achieve alignment across domains, and a weight allocation mechanism to suppress negative transfer.
IEEE-ASME TRANSACTIONS ON MECHATRONICS
(2022)
Article
Engineering, Industrial
Ahmed Maged et al.
Summary: This article proposes an approach using the prediction uncertainty information generated by Bayesian deep learning models to improve decision-making in fault detection in manufacturing. Experimental results demonstrate that the proposed scheme outperforms classical deep learning models.
JOURNAL OF MANUFACTURING SYSTEMS
(2022)
Article
Engineering, Industrial
Taotao Zhou et al.
Summary: This paper proposes a probabilistic Bayesian deep learning framework for fault diagnosis, which achieves trustworthy diagnosis by utilizing an uncertainty aware model. The framework accurately identifies faults and provides insights into uncertainty, avoiding erroneous decision-making, and demonstrates competitive results when considering unseen scenarios.
RELIABILITY ENGINEERING & SYSTEM SAFETY
(2022)
Article
Engineering, Industrial
Jingjie Luo et al.
Summary: In this paper, a modified deep subdomain adaptation network (MDSAN) is proposed for handling speed fluctuation in practical and challenging cross-domain diagnostic scenarios. The method utilizes a shared feature extraction module guided by a multi-headed self-attention mechanism to extract representative features, and a new trade-off factor is designed to improve convergence performance and optimization process.
JOURNAL OF MANUFACTURING SYSTEMS
(2022)
Article
Engineering, Multidisciplinary
Yiming Xiao et al.
Summary: Unsupervised cross-domain fault diagnosis research of rotating machinery is of significant importance for improving diagnostic performance. However, there are some existing issues, including the inability of convolutional neural network to extract discriminative information from vibration signals of different scales, the need to select extracted features to enhance informative features and suppress redundant features, and the potential problem of source and target domains being too close due to global domain adaptation methods.
Article
Engineering, Industrial
Te Han et al.
Summary: This article presents a novel OOD detection-assisted trustworthy machinery fault diagnosis approach, which integrates multiple deep neural networks and utilizes uncertainty analysis to enhance the reliability and safety of intelligent models, showing significant advantages in identifying OOD samples.
RELIABILITY ENGINEERING & SYSTEM SAFETY
(2022)
Article
Computer Science, Artificial Intelligence
Yasong Li et al.
Summary: This article proposes an interpretable deep learning model called VATN for rotary machine fault diagnosis. VATN improves upon the Transformer encoder to mine association relationships within signals and uses a sparse constraint to embed prior knowledge of fault types. Experimental studies demonstrate the effectiveness of VATN compared to other methods, and the heat map of attention weights illustrates the causal association between fault types and signal patterns.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Acoustics
Boyang Xue et al.
Summary: This paper proposes a Bayesian learning framework to address the overfitting and poor generalization issues of LSTM-RNNs and Transformers. Experimental results demonstrate that this framework significantly improves the performance of language models.
IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING
(2022)
Article
Engineering, Electrical & Electronic
Botao An et al.
Summary: This article proposes an interpretable neural network, named NISTA-Net, for mechanical fault diagnosis. The network is generated by unrolling the nested iterative soft thresholding algorithm (NISTA) and has a clear theoretical basis. A visualization method is also proposed to examine the meaningful features learned by NISTA-Net. Simulation and experiments demonstrate that NISTA-Net achieves superior performance in fault diagnosis compared to other advanced networks. A systematic approach is summarized to inspire further research in interpretable fault diagnosis networks.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2022)
Review
Computer Science, Artificial Intelligence
Moloud Abdar et al.
Summary: Uncertainty quantification (UQ) methods are essential in reducing uncertainties in optimization and decision making processes. Bayesian approximation and ensemble learning techniques are widely used types of UQ methods.
INFORMATION FUSION
(2021)
Article
Automation & Control Systems
Huan Wang et al.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2020)
Article
Automation & Control Systems
Zhibin Zhao et al.
Article
Engineering, Mechanical
Te Han et al.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2019)